Cognitive mobile radios in the longer term will require broadband small size antennas spanning the UHF and lower microwave mobile radio bands, typically 400 MHz to 3 GHz. The electrical characteristics, not least the bandwidth, of such an antenna will be dependent, amongst other things, on frequency of operation, ground plane size, profile thickness, overall volume occupied by the antenna structure and the immediate environment where the antenna is deployed. To meet the required characteristics the antenna must invariably be a reconfigurable structure that will service only relatively narrow parts of the whole band at a given time instant. Additionally it should be expected that the electrical characteristics, namely the radiation patterns, will vary across the whole spectrum of operation. This antenna therefore requires technology that allows hard switching and soft tuning mechanisms both in the radiating structure as well as in the feeding structure. The switching and tuning algorithms are either integrated in the general architecture of the cognitive radio or implemented as stand alone. These algorithms are expected to 'learn' in an efficient way how to tune the antennas that are under the influence of the environment.